There is a continuing need to improve the performance of computers to meet the needs of new and more sophisticated computing applications. Applications such as pattern classification, pattern association, associative memory functions, speech, and character recognition remain largely unamenable to solution or implementation by current computers as are many tasks that are readily and intuitively performed by humans and other biological organisms.
The desire to expand the frontiers of computer science has prompted consideration of the factors that contribute to the limitations of current computers. Silicon is at the heart of today's computer. The advances in computing power and speed over the years have largely been a consequence of better understanding the fundamental properties of silicon and harnessing those properties for practical effect. Initial progress was predicated on building basic electronic components such as transistors and diodes out of silicon and later progress followed from the development of integrated circuits. More recent advances represent a continuation of these trends and currently emphasize miniaturization and the integration of an ever larger number of microelectronic devices on a single chip. Smaller devices lead to higher memory storage densities, more highly integrated circuits and reduced interaction times between devices on the same chip.
Since future improvements in computing power and functionality are currently predicated on further improvements in silicon technology, there has been much recent discussion about the prognosis for continued miniaturization of silicon-based electronic devices. A growing consensus is emerging that believes that the computer industry is rapidly approaching the performance limits of silicon. The feature size in today's manufacturing technologies is 0.18 micron and it is expected that this can be reduced to about 0.10 micron in the future. Further decreases in feature size, however, are deemed problematic because sizes below about 0.10 micron lead to a change in the fundamental behavior of silicon. More specifically, as the dimensions of silicon devices decrease to tens of nanometers and below, silicon enters the quantum regime of behavior and no longer functions according to the classical physics that governs macroscopic objects. In the quantum regime, energy states are quantized rather than continuous and phenomena such as tunneling lead to delocalization of electrons across many devices. Consequences of tunneling include leakage of current as electrons escape from one device to neighboring devices and a loss of independence of devices as the state of one device influences the state of neighboring devices. In addition to fundamental changes in the behavior of silicon, further decreases in the dimensions of silicon devices also pose formidable technological challenges. New and costly innovations in fabrication methods such as photolithography will be needed to achieve smaller feature sizes.
One strategy for advancing the capabilities of computers is to identify materials other than silicon that can be used as the active medium in data processing and/or storage applications. Such alternative computing media could be used independent of or in combination with silicon to form the basis of a new computing industry that seeks to offer better performance and more convenient manufacturing than is possible with silicon.
The instant inventors have recently proposed the use of chalcogenide phase change materials as an active material for the processing and storage of data. In U.S. Pat. No. 6,671,710 (the '710 patent), the disclosure of which is hereby incorporated by reference herein, Ovshinsky et al. describe a principle of operation of phase change materials in computing applications. Phase change materials can not only operate in the binary mode characteristic of conventional silicon computers, but also offer opportunities for the non-binary storage and processing of data. Non-binary storage provides for high information storage densities, while non-binary processing provides for increased parallelness of operation. The '710 patent also describes representative algorithms that utilize a non-binary computing medium for mathematical operations such as addition, subtraction, multiplication and division. U.S. Pat. No. 6,714,954 (the '954 patent) by Ovshinsky et al., the disclosure of which is hereby incorporated by reference herein, describes farther mathematical operations based on a phase change computing medium, including factoring, modular arithmetic and parallel operation.
In U.S. Pat. No. 6,999,953 (the '953 patent), the disclosure of which is hereby incorporated by reference herein, Ovshinsky considers the architecture of computing systems based on devices utilizing a phase change material as the active computing medium. More specifically, Ovshinsky considers networks of phase change computing devices and demonstrates functionality tat closely parallels that of biological neural networks. Important features of this functionality include the accumulative response of phase change computing devices to input signals from a variety of sources, an ability to weight the input signals and a stable, reproducible material transformation that mimics the firing of a biological neuron. This functionality enables a new concept in intelligent computing that features learning, adaptability, and plasticity.
In U.S. Pat. Nos. 6,967,344 (the '344 patent); U.S. Pat. No. 6,969,867 (the '867 patent); and U.S. Pat. No. 7,227,170 (the '170 patent), the disclosures of which are hereby incorporated by reference herein, Ovshinsky et al. further develop the notion of phase change computing by discussing additional computing and storage devices. The '344 patent discusses a multi-terminal phase change device where a control signal provided at one electrical terminal modulates the current, threshold voltage or signal transmitted between other electrical terminals through the injection of charge carriers. The '867 patent describes a related multi-terminal device that utilizes a field effect terminal to modulate the current, threshold voltage or signal transmitted between other terminals. The devices described in the '344 and '867 patents may be configured to provide a functionality analogous to that of the transistor that is so vital to silicon based computers. The '170 patent presents a multiple bit storage device having multiple terminals that utilizes a phase change material.
The foregoing work by Ovshinsky et al. provides a concept, operating principles and some basic devices to enable a computing paradigm based in whole or in part on chalcogenide or other phase change materials. In order to further the realization of chalcogenide computing as a viable complement or alternative to silicon-based technologies, it is desirable to expand the range of devices and functionality available from chalcogenide phase change materials. Of greatest interest are devices and systems capable of performing processing, storage or memory, and logic functions.